1. 26 Nov, 2010 14 commits
  2. 23 Nov, 2010 1 commit
    • Rabin Vincent's avatar
      perf symbols: Remove incorrect open-coded container_of() · 02a9d037
      Rabin Vincent authored
      At least on ARM, padding is inserted between rb_node and sym in struct
      symbol_name_rb_node, causing "((void *)sym) - sizeof(struct rb_node)" to
      point inside rb_node rather than to the symbol_name_rb_node.  Fix this
      by converting the code to use container_of().
      
      Cc: Ian Munsie <imunsie@au1.ibm.com>
      Cc: Ingo Molnar <mingo@elte.hu>
      Cc: Ming Lei <tom.leiming@gmail.com>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Thomas Gleixner <tglx@linutronix.de>
      Cc: Tom Zanussi <tzanussi@gmail.com>
      LKML-Reference: <20101123163106.GA25677@debian>
      Signed-off-by: default avatarRabin Vincent <rabin@rab.in>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      02a9d037
  3. 22 Nov, 2010 1 commit
    • Arnaldo Carvalho de Melo's avatar
      perf record: Handle restrictive permissions in /proc/{kallsyms,modules} · c1a3a4b9
      Arnaldo Carvalho de Melo authored
      The 59365d13 commit, even being reverted by 33e0d57f, showed a non robust
      behavior in 'perf record': it really should just warn the user that some
      functionality will not be available.
      
      The new behavior then becomes:
      
      	[acme@felicio linux]$ ls -la /proc/{kallsyms,modules}
      	-r-------- 1 root root 0 Nov 22 12:19 /proc/kallsyms
      	-r-------- 1 root root 0 Nov 22 12:19 /proc/modules
      	[acme@felicio linux]$ perf record ls -R > /dev/null
      	Couldn't record kernel reference relocation symbol
      	Symbol resolution may be skewed if relocation was used (e.g. kexec).
      	Check /proc/kallsyms permission or run as root.
      	[ perf record: Woken up 1 times to write data ]
      	[ perf record: Captured and wrote 0.004 MB perf.data (~161 samples) ]
      	[acme@felicio linux]$ perf report --stdio
      	[kernel.kallsyms] with build id 77b05e00e64e4de1c9347d83879779b540d69f00 not found, continuing without symbols
      	# Events: 98  cycles
      	#
      	# Overhead  Command    Shared Object                Symbol
      	# ........  .......  ...............  ....................
      	#
      	    48.26%       ls  [kernel]         [k] ffffffff8102b92b
      	    22.49%       ls  libc-2.12.90.so  [.] __strlen_sse2
      	     8.35%       ls  libc-2.12.90.so  [.] __GI___strcoll_l
      	     8.17%       ls  ls               [.]            11580
      	     3.35%       ls  libc-2.12.90.so  [.] _IO_new_file_xsputn
      	     3.33%       ls  libc-2.12.90.so  [.] _int_malloc
      	     1.88%       ls  libc-2.12.90.so  [.] _int_free
      	     0.84%       ls  libc-2.12.90.so  [.] malloc_consolidate
      	     0.84%       ls  libc-2.12.90.so  [.] __readdir64
      	     0.83%       ls  ls               [.] strlen@plt
      	     0.83%       ls  libc-2.12.90.so  [.] __GI_fwrite_unlocked
      	     0.83%       ls  libc-2.12.90.so  [.] __memcpy_sse2
      
      	#
      	# (For a higher level overview, try: perf report --sort comm,dso)
      	#
      [acme@felicio linux]$
      
      It still has the build-ids for DSOs in the maps with hits:
      
      [acme@felicio linux]$ perf buildid-list
      77b05e00e64e4de1c9347d83879779b540d69f00 [kernel.kallsyms]
      09c4a431a4a8b648fcfc2c2bdda70f56050ddff1 /bin/ls
      af75ea9ad951d25e0f038901a11b3846dccb29a4 /lib64/libc-2.12.90.so
      [acme@felicio linux]$
      
      That can be used in another machine to resolve kernel symbols.
      
      Cc: Eugene Teo <eugeneteo@kernel.org>
      Cc: Frederic Weisbecker <fweisbec@gmail.com>
      Cc: Ingo Molnar <mingo@elte.hu>
      Cc: Jesper Juhl <jj@chaosbits.net>
      Cc: Marcus Meissner <meissner@suse.de>
      Cc: Mike Galbraith <efault@gmx.de>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Peter Zijlstra <peterz@infradead.org>
      Cc: Sarah Sharp <sarah.a.sharp@linux.intel.com>
      Cc: Stephane Eranian <eranian@google.com>
      Cc: Tejun Heo <tj@kernel.org>
      Cc: Tom Zanussi <tzanussi@gmail.com>
      LKML-Reference: <new-submission>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      c1a3a4b9
  4. 21 Nov, 2010 1 commit
  5. 20 Nov, 2010 4 commits
    • Corey Ashford's avatar
      perf stat: Change and clean up sys_perf_event_open error handling · d9cf837e
      Corey Ashford authored
      This patch makes several changes to "perf stat":
      
      - "perf stat" will no longer go ahead and run the application when one or
      more of the specified events could not be opened.
      - Use error() and die() instead of pr_err() so that the output is more
      consistent with "perf top" and "perf record".
      - Handle permission errors in a more robust way, and in a similar way to
      "perf record" and "perf top".
      
      In addition, the sys_perf_event_open() error handling of "perf top" and "perf
      record" is made more consistent and adds the following phrase when an event
      doesn't open (with something ther than an access or permission error):
      
      "/bin/dmesg may provide additional information."
      
      This is added because kernel code doesn't have a good way of expressing
      detailed errors to user space, so its only avenue is to use printk's.  However,
      many users may not think of looking at dmesg to find out why an event is being
      rejected.
      
      Cc: Frederic Weisbecker <fweisbec@gmail.com>
      Cc: Ingo Molnar <mingo@elte.hu>
      Cc: Peter Zijlstra <fweisbec@gmail.com>
      Cc: Ian Munsie <ianmunsi@au1.ibm.com>
      Cc: Michael Ellerman <michaele@au1.ibm.com>
      LKML-Reference: <1290217044-26293-1-git-send-email-cjashfor@linux.vnet.ibm.com>
      Signed-off-by: default avatarCorey Ashford <cjashfor@linux.vnet.ibm.com>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      d9cf837e
    • Linus Torvalds's avatar
      Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4 · b86db474
      Linus Torvalds authored
      * 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
        ext4: Add EXT4_IOC_TRIM ioctl to handle batched discard
        fs: Do not dispatch FITRIM through separate super_operation
        ext4: ext4_fill_super shouldn't return 0 on corruption
        jbd2: fix /proc/fs/jbd2/<dev> when using an external journal
        ext4: missing unlock in ext4_clear_request_list()
        ext4: fix setting random pages PageUptodate
      b86db474
    • Lukas Czerner's avatar
      ext4: Add EXT4_IOC_TRIM ioctl to handle batched discard · e681c047
      Lukas Czerner authored
      Filesystem independent ioctl was rejected as not common enough to be in
      core vfs ioctl. Since we still need to access to this functionality this
      commit adds ext4 specific ioctl EXT4_IOC_TRIM to dispatch
      ext4_trim_fs().
      
      It takes fstrim_range structure as an argument. fstrim_range is definec in
      the include/linux/fs.h and its definition is as follows.
      
      struct fstrim_range {
      	__u64 start;
      	__u64 len;
      	__u64 minlen;
      }
      
      start	- first Byte to trim
      len	- number of Bytes to trim from start
      minlen	- minimum extent length to trim, free extents shorter than this
        number of Bytes will be ignored. This will be rounded up to fs
        block size.
      
      After the FITRIM is done, the number of actually discarded Bytes is stored
      in fstrim_range.len to give the user better insight on how much storage
      space has been really released for wear-leveling.
      Signed-off-by: default avatarLukas Czerner <lczerner@redhat.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      e681c047
    • Lukas Czerner's avatar
      fs: Do not dispatch FITRIM through separate super_operation · 93bb41f4
      Lukas Czerner authored
      There was concern that FITRIM ioctl is not common enough to be included
      in core vfs ioctl, as Christoph Hellwig pointed out there's no real point
      in dispatching this out to a separate vector instead of just through
      ->ioctl.
      
      So this commit removes ioctl_fstrim() from vfs ioctl and trim_fs
      from super_operation structure.
      Signed-off-by: default avatarLukas Czerner <lczerner@redhat.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      93bb41f4
  6. 19 Nov, 2010 18 commits
    • Linus Torvalds's avatar
      Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client · 76db8ac4
      Linus Torvalds authored
      * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
        ceph: fix readdir EOVERFLOW on 32-bit archs
        ceph: fix frag offset for non-leftmost frags
        ceph: fix dangling pointer
        ceph: explicitly specify page alignment in network messages
        ceph: make page alignment explicit in osd interface
        ceph: fix comment, remove extraneous args
        ceph: fix update of ctime from MDS
        ceph: fix version check on racing inode updates
        ceph: fix uid/gid on resent mds requests
        ceph: fix rdcache_gen usage and invalidate
        ceph: re-request max_size if cap auth changes
        ceph: only let auth caps update max_size
        ceph: fix open for write on clustered mds
        ceph: fix bad pointer dereference in ceph_fill_trace
        ceph: fix small seq message skipping
        Revert "ceph: update issue_seq on cap grant"
      76db8ac4
    • Linus Torvalds's avatar
      Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6 · caf83945
      Linus Torvalds authored
      * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (31 commits)
        net: fix kernel-doc for sk_filter_rcu_release
        be2net: Fix to avoid firmware update when interface is not open.
        netfilter: fix IP_VS dependencies
        net: irda: irttp: sync error paths of data- and udata-requests
        ipv6: Expose reachable and retrans timer values as msecs
        ipv6: Expose IFLA_PROTINFO timer values in msecs instead of jiffies
        3c59x: fix build failure on !CONFIG_PCI
        ipg.c: remove id [SUNDANCE, 0x1021]
        net: caif: spi: fix potential NULL dereference
        ath9k_htc: Avoid setting QoS control for non-QoS frames
        net: zero kobject in rx_queue_release
        net: Fix duplicate volatile warning.
        MAINTAINERS: Add stmmac maintainer
        bonding: fix a race in IGMP handling
        cfg80211: fix can_beacon_sec_chan, reenable HT40
        gianfar: fix signedness issue
        net: bnx2x: fix error value sign
        8139cp: fix checksum broken
        r8169: fix checksum broken
        rds: Integer overflow in RDS cmsg handling
        ...
      caf83945
    • Linus Torvalds's avatar
    • Linus Torvalds's avatar
      Revert "kernel: make /proc/kallsyms mode 400 to reduce ease of attacking" · 33e0d57f
      Linus Torvalds authored
      This reverts commit 59365d13.
      
      It turns out that this can break certain existing user land setups.
      Quoth Sarah Sharp:
      
       "On Wednesday, I updated my branch to commit 460781b5 from linus' tree,
        and my box would not boot.  klogd segfaulted, which stalled the whole
        system.
      
        At first I thought it actually hung the box, but it continued booting
        after 5 minutes, and I was able to log in.  It dropped back to the
        text console instead of the graphical bootup display for that period
        of time.  dmesg surprisingly still works.  I've bisected the problem
        down to this commit (commit 59365d13)
      
        The box is running klogd 1.5.5ubuntu3 (from Jaunty).  Yes, I know
        that's old.  I read the bit in the commit about changing the
        permissions of kallsyms after boot, but if I can't boot that doesn't
        help."
      
      So let's just keep the old default, and encourage distributions to do
      the "chmod -r /proc/kallsyms" in their bootup scripts.  This is not
      worth a kernel option to change default behavior, since it's so easily
      done in user space.
      Reported-and-bisected-by: default avatarSarah Sharp <sarah.a.sharp@linux.intel.com>
      Cc: Marcus Meissner <meissner@suse.de>
      Cc: Tejun Heo <tj@kernel.org>
      Cc: Eugene Teo <eugeneteo@kernel.org>
      Cc: Jesper Juhl <jj@chaosbits.net>
      Signed-off-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      33e0d57f
    • Arnaldo Carvalho de Melo's avatar
      perf tools: Change my maintainer address · 4aafd3f7
      Arnaldo Carvalho de Melo authored
      Also remove old snail mail address from CREDITS, moved years ago.
      
      LKML-Reference: <new-submission>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      4aafd3f7
    • Robert Morell's avatar
      perf tools: Remove hardcoded include paths for elfutils · a7112397
      Robert Morell authored
      This change removes the use of hardcoded absolute "/usr/include/elfutils" paths
      from the perf build.  The problem with hardcoded paths is that it prevents them
      from being overridden by $prefix or by -I in CFLAGS (e.g., for cross-compiling
      purposes).
      
      Instead, just include the "elfutils/" subdirectory as a relative path when
      files are needed from that directory.
      
      Tested by building perf:
      - Cross-compiled for ARM on x86_64
      - Built natively on x86_64
      - Built on x86_64 with /usr/include/elfutils moved to another location
        and manually included in CFLAGS
      Acked-by: default avatarMasami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
      Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Ingo Molnar <mingo@elte.hu>
      Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
      LKML-Reference: <1289945793-31441-1-git-send-email-rmorell@nvidia.com>
      Signed-off-by: default avatarRobert Morell <rmorell@nvidia.com>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      a7112397
    • Linus Torvalds's avatar
      Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input · 864ee6cb
      Linus Torvalds authored
      * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input:
        Input: fix typo in keycode validation supporting large scancodes
        Input: aiptek - tighten up permissions on sysfs attributes
        Input: sysrq - pass along lone Alt + SysRq
      864ee6cb
    • Linus Torvalds's avatar
      Merge branch 'drm-intel-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/ickle/drm-intel · 973d168d
      Linus Torvalds authored
      * 'drm-intel-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/ickle/drm-intel:
        drm/i915: Disable FBC on Ironlake to save 1W
        drm/i915: Take advantage of auto-polling CRT hotplug detection on PCH hardware
        drm/i915/crt: Introduce struct intel_crt
        drm/i915: Do not hold mutex when faulting in user addresses
        drm: radeon: fix error value sign
        drm/radeon/kms: fix and unify tiled buffer alignment checking for r6xx/7xx
        drm/i915: Retire any pending operations on the old scanout when switching
        drm/i915: Fix I2C adapter registration
      973d168d
    • Linus Torvalds's avatar
      Merge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6 · 764bc569
      Linus Torvalds authored
      * 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6: (40 commits)
        drm/radeon/kms: i2c s/sprintf/snprintf/g for safety
        drm/radeon/kms: fix i2c pad masks on rs4xx
        drm/ttm: Fix up a theoretical deadlock
        drm/radeon/kms: fix tiling info on evergreen
        drm/radeon/kms: fix alignment when allocating buffers
        drm/vmwgfx: Fix up an error path during bo creation
        drm/radeon/kms: register an i2c adapter name for the dp aux bus
        drm/radeon/kms/atom: add proper external encoders support
        drm/radeon/kms/atom: cleanup and unify DVO handling
        drm/radeon/kms: properly power up/down the eDP panel as needed (v4)
        drm/radeon/kms/atom: set sane defaults in atombios_get_encoder_mode()
        drm/radeon/kms: turn the backlight off explicitly for dpms
        drm/radeon/kms: fix typo in r600 cs checker
        drm: radeon: fix error value sign
        drm/radeon/kms: fix and unify tiled buffer alignment checking for r6xx/7xx
        nouveau: Acknowledge HPD irq in handler, not bottom half
        drm/nouveau: Fix a few confusions between "chipset" and "card_type".
        drm/nouveau: don't expose backlight control when available through ACPI
        drm/nouveau/pm: improve memtiming mappings
        drm/nouveau: Make PCIE GART size depend on the available RAMIN space.
        ...
      764bc569
    • Stephane Eranian's avatar
      perf stat: Add no-aggregation mode to -a · f5b4a9c3
      Stephane Eranian authored
      This patch adds a new -A option to perf stat. If specified then perf stat does
      not aggregate counts across all monitored CPUs in system-wide mode, i.e., when
      using -a. This option is not supported in per-thread mode.
      
      Being able to get a per-cpu breakdown is useful to detect imbalances between
      CPUs when running a uniform workload than spans all monitored CPUs.
      
      The second version corrects the missing cpumap[] support, so that it works when
      the -C option is used.
      
      The third version fixes a missing cpumap[] in print_counter() and removes a
      stray patch in builtin-trace.c.
      
      Examples on a 4-way system:
      
      # perf stat -a   -e cycles,instructions -- sleep 1
       Performance counter stats for 'sleep 1':
               9592808135  cycles
               3490380006  instructions             #      0.364 IPC
              1.001584632  seconds time elapsed
      
      # perf stat -a -A -e cycles,instructions -- sleep 1
       Performance counter stats for 'sleep 1':
      CPU0            2398163767  cycles
      CPU1            2398180817  cycles
      CPU2            2398217115  cycles
      CPU3            2398247483  cycles
      CPU0             872282046  instructions             #      0.364 IPC
      CPU1             873481776  instructions             #      0.364 IPC
      CPU2             872638127  instructions             #      0.364 IPC
      CPU3             872437789  instructions             #      0.364 IPC
              1.001556052  seconds time elapsed
      
      Cc: David S. Miller <davem@davemloft.net>
      Cc: Frederic Weisbecker <fweisbec@gmail.com>
      Cc: Ingo Molnar <mingo@elte.hu>
      Cc: Paul Mackerras <paulus@samba.org>
      Cc: Peter Zijlstra <peterz@infradead.org>
      Cc: Robert Richter <robert.richter@amd.com>
      LKML-Reference: <4ce257b5.1e07e30a.7b6b.3aa9@mx.google.com>
      Signed-off-by: default avatarStephane Eranian <eranian@google.com>
      Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
      f5b4a9c3
    • Tejun Heo's avatar
      sata_via: apply magic FIFO fix to vt6420 too · b1353e4f
      Tejun Heo authored
      vt6420 has the same FIFO overflow problem as vt6421 when combined with
      certain devices.  This patch applies the magic fix to vt6420 too.
      Signed-off-by: default avatarTejun Heo <tj@kernel.org>
      Reported-by: default avatarMartin Qvist <q@maq.dk>
      Reported-by: default avatarPeter Zijlstra <peterz@infradead.org>
      Cc: Joseph Chan <JosephChan@via.com.tw>
      Cc: stable@kernel.org
      Signed-off-by: default avatarJeff Garzik <jgarzik@redhat.com>
      b1353e4f
    • Randy Dunlap's avatar
      net: fix kernel-doc for sk_filter_rcu_release · 0302b862
      Randy Dunlap authored
      Fix kernel-doc warning for sk_filter_rcu_release():
      
      Warning(net/core/filter.c:586): missing initial short description on line:
       * 	sk_filter_rcu_release: Release a socket filter by rcu_head
      Signed-off-by: default avatarRandy Dunlap <randy.dunlap@oracle.com>
      Cc:	"David S. Miller" <davem@davemloft.net>
      Cc:	netdev@vger.kernel.org
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      0302b862
    • Sarveshwar Bandi's avatar
      be2net: Fix to avoid firmware update when interface is not open. · d9efd2af
      Sarveshwar Bandi authored
      Since interrupts are enabled only when open is called on the interface,
      Attempting a firmware update operation when interface is down could lead to
      partial success or failure of operation. This fix fails the request if
      netif_running is false.
      Signed-off-by: default avatarSarveshwar Bandi <Sarveshwar.Bandi@emulex.com>
      Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      d9efd2af
    • Darrick J. Wong's avatar
      ext4: ext4_fill_super shouldn't return 0 on corruption · 5a9ae68a
      Darrick J. Wong authored
      At the start of ext4_fill_super, ret is set to -EINVAL, and any failure path
      out of that function returns ret.  However, the generic_check_addressable
      clause sets ret = 0 (if it passes), which means that a subsequent failure (e.g.
      a group checksum error) returns 0 even though the mount should fail.  This
      causes vfs_kern_mount in turn to think that the mount succeeded, leading to an
      oops.
      
      A simple fix is to avoid using ret for the generic_check_addressable check,
      which was last changed in commit 30ca22c7.
      Signed-off-by: default avatarDarrick J. Wong <djwong@us.ibm.com>
      Signed-off-by: default avatar"Theodore Ts'o" <tytso@mit.edu>
      5a9ae68a
    • Alex Shi's avatar
      drm/i915: Disable FBC on Ironlake to save 1W · 16c59ef3
      Alex Shi authored
      Frame buffer compression is broken on Ironlake due to buggy hardware.
      Currently it is disabled through chicken bits, but it still consumes
      over 1W more than if we simply never attempt to enable the FBC code
      paths.
      Signed-off-by: default avatarAlex Shi <alex.shi@intel.com>
      Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
      Cc: stable@kernel.org
      16c59ef3
    • Keith Packard's avatar
      drm/i915: Take advantage of auto-polling CRT hotplug detection on PCH hardware · e7dbb2f2
      Keith Packard authored
      Both IBX and CPT have an automatic hotplug detection mode which appears to work reliably enough
      that we can dispense with the manual force hotplug trigger stuff. This means that
      hotplug detection is as simple as reading the current hotplug register values.
      
      The first time the hotplug detection is activated, the code synchronously waits for a hotplug
      sequence in case the hardware hasn't bothered to do a detection cycle since being initialized.
      Signed-off-by: default avatarKeith Packard <keithp@keithp.com>
      Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
      e7dbb2f2
    • Chris Wilson's avatar
      drm/i915/crt: Introduce struct intel_crt · c9a1c4cd
      Chris Wilson authored
      We will use this structure in future patches to store CRT specific
      information on the encoder.
      
      Split out and tweaked from a patch by Keith Packard.
      Signed-off-by: default avatarKeith Packard <keithp@kithp.com>
      Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
      c9a1c4cd
    • Chris Wilson's avatar
      drm/i915: Do not hold mutex when faulting in user addresses · 51311d0a
      Chris Wilson authored
      Linus Torvalds found that it was rather trivial to trigger a system
      freeze:
      
        In fact, with lockdep, I don't even need to do the sysrq-d thing: it
        shows the bug as it happens. It's the X server taking the same lock
        recursively.
      
        Here's the problem:
      
          =============================================
          [ INFO: possible recursive locking detected ]
          2.6.37-rc2-00012-gbdbd01ac #7
          ---------------------------------------------
          Xorg/2816 is trying to acquire lock:
           (&dev->struct_mutex){+.+.+.}, at: [<ffffffff812c626c>] i915_gem_fault+0x50/0x17e
      
          but task is already holding lock:
           (&dev->struct_mutex){+.+.+.}, at: [<ffffffff812c403b>] i915_mutex_lock_interruptible+0x28/0x4a
      
          other info that might help us debug this:
          2 locks held by Xorg/2816:
           #0:  (&dev->struct_mutex){+.+.+.}, at: [<ffffffff812c403b>] i915_mutex_lock_interruptible+0x28/0x4a
           #1:  (&mm->mmap_sem){++++++}, at: [<ffffffff81022d4f>] page_fault+0x156/0x37b
      
      This recursion was introduced by rearranging the locking to avoid the
      double locking on the fast path (4f27b5d and fbd5a26d) and the
      introduction of the prefault to encourage the fast paths (b5e4f2b). In
      order to undo the problem, we rearrange the code to perform the access
      validation upfront, attempt to prefault and then fight for control of the
      mutex.  the best case scenario where the mutex is uncontended the
      prefaulting is not wasted.
      Reported-and-tested-by: default avatarLinus Torvalds <torvalds@linux-foundation.org>
      Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
      51311d0a
  7. 18 Nov, 2010 1 commit